rename filterTextFlags to disableLCD

rename filterTextFlags to disableLCD

Under the hood, add SkPixelGeometry to the CreateInfo for new devices, allowing them to see their geometry (SkDeviceProperties) up front, rather than having it changed later.

The only exception is for devices that are used on the root-layer, where we don't see the device until after the fact (at least as long as we allow clients to attach a device to a canvas externally).

We also filter the geometry when we're creating a layer, so we can disable LCD text automatically if the layer is not marked as opaque.

NOTRY=True
-- gammatext flake?

Committed: https://skia.googlesource.com/skia/+/b2db898573e3cdcc8234eebf51961bfc4977ebbc

[reed1, 2014-11-12 20:32:39]

reed@google.com changed reviewers:
+ bsalomon@google.com, fmalita@google.com

[f(malita), 2014-11-12 20:49:13]

fmalita@chromium.org changed reviewers:
+ fmalita@chromium.org

[f(malita), 2014-11-12 20:49:14]

Mostly nits.

https://codereview.chromium.org/719253002/diff/1/include/core/SkBitmapDevice.h
File include/core/SkBitmapDevice.h (right):

https://codereview.chromium.org/719253002/diff/1/include/core/SkBitmapDevice....
include/core/SkBitmapDevice.h:41: bool onDisableLCD(const SkPaint&) const
SK_OVERRIDE;
virtual

https://codereview.chromium.org/719253002/diff/1/include/core/SkDevice.h
File include/core/SkDevice.h (right):

https://codereview.chromium.org/719253002/diff/1/include/core/SkDevice.h#newc...
include/core/SkDevice.h:121: bool disableLCD(const SkPaint&) const;
This sounds imperative. Maybe canHandleLCD or supportsLCD?

https://codereview.chromium.org/719253002/diff/1/src/core/SkBitmapDevice.cpp
File src/core/SkBitmapDevice.cpp (right):

https://codereview.chromium.org/719253002/diff/1/src/core/SkBitmapDevice.cpp#...
src/core/SkBitmapDevice.cpp:394: return true;
Maybe inline the condition in a single return statement?

https://codereview.chromium.org/719253002/diff/1/src/core/SkCanvas.cpp
File src/core/SkCanvas.cpp (right):

https://codereview.chromium.org/719253002/diff/1/src/core/SkCanvas.cpp#newcod...
src/core/SkCanvas.cpp:2058: if (device->disableLCD(paint)) {
This approach requires all clients to munge the LCD-related flags & duplicate
this logic (currently this is the only place, but drawtextblob will also need to
do it). For this reason alone I think I would prefer the filterFlags() variant
where munging is encapsulated and clients don't really need to know what LCD is
and how to turn it off.

https://codereview.chromium.org/719253002/diff/1/src/gpu/SkGpuDevice.h
File src/gpu/SkGpuDevice.h (right):

https://codereview.chromium.org/719253002/diff/1/src/gpu/SkGpuDevice.h#newcod...
src/gpu/SkGpuDevice.h:125: bool onDisableLCD(const SkPaint&) const SK_OVERRIDE;
virtual

[bsalomon, 2014-11-13 14:37:42]

bsalomon@google.com changed reviewers:
+ jvanverth@google.com

[bsalomon, 2014-11-13 14:37:42]

I defer to Jim

[reed1, 2014-11-13 20:03:38]

The CQ bit was checked by reed@google.com

[commit-bot: I haz the power, 2014-11-13 20:03:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/719253002/40001

[commit-bot: I haz the power, 2014-11-13 20:03:58]

Note for Reviewers:
The CQ is waiting for an approval.
If you believe that the CL is not ready yet, or if you would like to L-G-T-M
with comments then please uncheck the CQ checkbox.

Waiting for LGTM from valid reviewer(s) till 2014-11-14 02:03 UTC

[f(malita), 2014-11-13 20:16:55]

The CQ bit was unchecked by fmalita@chromium.org

[f(malita), 2014-11-13 20:17:38]

LGTM w/ name change.

[reed1, 2014-11-13 20:19:39]

The CQ bit was checked by reed@google.com

[commit-bot: I haz the power, 2014-11-13 20:20:28]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/719253002/60001

[reed1, 2014-11-13 20:38:15]

The CQ bit was unchecked by reed@google.com

[reed1, 2014-11-13 20:38:24]

The CQ bit was checked by reed@google.com

[commit-bot: I haz the power, 2014-11-13 20:38:44]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/719253002/60001

[commit-bot: I haz the power, 2014-11-13 20:41:09]

Committed patchset #4 (id:60001) as
https://skia.googlesource.com/skia/+/b2db898573e3cdcc8234eebf51961bfc4977ebbc

[brucedawson, 2014-12-10 01:37:05]

Just thought I'd point out a perf-win from this change that I didn't see
mentioned in the description. Prior to this change
createCompatibleDeviceForSaveLayer was getting called frequently when typing
into gmail and the large allocations this triggered were causing a noticeable
amount of overhead. I went to investigate and found that the allocations were
already gone in canary, presumably due to this change.

So, thanks!